(35e) Multi-State Monte Carlo Optimization of Electrostatic Domain Differences for Self-Assembling Proteins

Protein fibers and hydrogels are becoming more prevalent in biomedicine for a variety of applications. We have previously synthesized an upper critical solution temperature-type (UCST) hydrogel, Q: a homopentameric coiled-coil protein hydrogel capable of demonstrating a UCST and encapsulating small hydrophobic molecules. However, Q displays a performance gap in its ability to undergo its UCST and self-assemble at physiological conditions. The rationally designed Q protein undergoes hierarchical self-assembly into nanofibers mediated by intermolecular associations between coiled-coil pentamers. Using this rationale, electrostatic potential of surface patches and thermostability are used in a multimodal Monte Carlo search algorithm to make selective mutations to generate Rationally Randomized Dielectric Domain (R2D2) proteins where surface charge is redistributed for favorable proto-fibril assembly. Microscopy and structural analysis reveal enhanced gelation kinetics and increased stability at physiological conditions. The resulting R2D2 proteins are a significant improvement in thermostability and sol-gel transition of a single coiled-coil domain capable of sustained release, by demonstrating a UCST-type behavior.